Extended Data Fig. 4: Quantification and quality control of RNA subcellular kinetic parameters.

a, Correlated relation between cell volume (in voxels) and single-cell reads, indicating the influence of transcript number by cell volume. b, Schematics showing the biased single-cell RNA read counts with varying physical cell volumes. c, Mathematical models for estimating RNA kinetic parameters (α,β and λ) and the detailed workflow of calculation and fitting procedure. Note: X(t) = single-cell RNA concentration; N(t) = nuclear RNA concentration; C(t) = cytoplasmic RNA concentration. d, Changes in the natural log of X(t) across timepoints of genes with R² = 0.99-1.0 (top) and R² = 0.46-0.53 (down). Representative genes are shown. The estimated β values for all genes were filtered with a threshold of R² > 0.5 as quality control. e-f, Density plots showing the distribution of R² when fitting degradation constant by a first-order ODE using original (e) and temporally-shuffled (f) pseudobulk expression vectors. g-h, scatterplots showing Pearson’s correlation of RNA synthesis (RNA/hour, g) and degradation (1/hr, h) estimation between this study and single-cell scEU-seq datasets after filtering the genes based on fitting the degradation parameters using R² > = 0.5, 0.8 and 0.9, respectively. Units of α is RNA concentration/hr, where RNA concentration is defined by copies/voxel (voxel = 200 nm x 200 nm x 350 nm). Units of β is hr⁻¹. i, Left: schematics of pulse-chase designs utilized in scEU-seq and scNT-seq; right: correlation of degradation between scEU-seq and scNT-seq after filtering genes based on fitting the degradation parameters using R² > = 0.9. j, Histogram of estimated λ (nuclear export) values for all genes. k, The distribution of single cell-averaged DR values for all 991 genes across 0-6 hrs chase timepoints. n = 991 genes for each timepoint. Boxplots are defined in terms of mean (center line), 25-75% percentile (bounds of box), lower and upper quartile (whiskers). l, Histogram of estimated γ (cytoplasmic translocation) values for all genes. Blue dashed line separates the genes with γ > 0 and γ < 0, which indicates the opposite direction of observed translocation. m, Left, 19 genes with γ < 0 (R² > 0.5) were strongly enriched in secreted and organellar proteins. Middle, time-lapsed DR values of representative genes. n, Schematics showing the observed inward direction of RNA translocation of genes with γ < 0. Two potential mechanisms are shown: 1). RNA translocation (active or passive); 2). Local RNA degradation.